light traction: fuel cells - elsevier.com · 2013. 12. 18. · light traction applications, it is...

Light Traction: Fuel Cells

Z Qi, Plug Power Inc., Latham, NY, USA

& 2009 Elsevier B.V. All rights reserved.

Introduction

As a proton-exchange membrane fuel cell (PEMFC) canbe started instantly at ambient temperatures and canwork with air as oxidant without carbon dioxide prob-lems, it is so far the most viable fuel cell (FC) system thathas the potential to replace internal combustion engines(ICEs) and batteries for transportation applications topower cars, buses, and personal electric vehicles (PEVs).This article focuses on the application of this technologyfor light traction vehicles, such as scooters, bicycles,forklifts, wheelchairs, and tour carts.

Fuel Cell Systems

There is no doubt that a vehicle should be able to startinstantly regardless of the powering technology used.Because onboard hydrocarbon reforming takes too muchtime to generate hydrogen, an FC that powers a vehiclehas to use either hydrogen or liquid fuels directly. Forlight traction applications, it is also desirable that the FCsystem is simple, compact, and cost-effective.

Hydrogen–Air Systems

Figure 1 depicts a hydrogen–air FC system. Basically, anFC stack receives hydrogen from a storage tank andoxygen from air through a fan, a blower, or a pump. Thedirect current (DC) electrical power generated by the FCis used to power an electrical motor that drives a vehicle.

The motor is preferred to be one that can operate with awide voltage range so that a DC–DC converter does notneed to be used, thereby reducing cost and increasingsystem efficiency. In addition, in-wheel (or wheel-hub)motor configurations where an electrical motor is locatedin the wheel should be considered in order to save spaceand to eliminate freewheeling. A microprocessor can beused to control the system operation.

In order to reduce the size and weight of the FC stackand to meet acceleration and hill climbing demands, it ishelpful to include either a battery or a capacitor in thesystem. A battery or a capacitor can provide burst powerneeded during acceleration and hill climbing. In addition,they can store waste energy through regenerative brak-ing, where the momentum of the wheels during brakingdrives the electric motor to run in reverse so that thelatter becomes a generator that produces electricity tocharge the battery or the capacitor. When the vehicle isunder cruise operation, the battery or the capacitor isrecharged by the FC. Another FC/battery configuration,not shown here, is to use the battery only to drive themotor while the FC is just for charging the battery, i.e.,the FC acts as a range extender. In either case, becausethe battery is frequently charged by the FC, the deepcharge/discharge cycles that normally shorten batterylifetime can be avoided. A supercapacitor has lower in-ternal resistance, more charging/discharging cycles andless efficiency loss during cycles, and requires lowermaintenance than a battery. Moreover, because a super-capacitor stores and provides electricity in response tothe fluctuations of the FC output, it does not require aconverter for voltage regulation.

The amount of hydrogen stored in a storage tankdetermines how long a vehicle can operate before thetank needs to be refilled or replaced. Presently, the mostviable ways of storing hydrogen are using either metalhydrides or high-pressure bottles. The former, wherehydrogen is absorbed in the intermetal lattices of a solid,often stores hydrogen at no more than 1 MPa, whereasthe latter can have hydrogen compressed to about35 MPa. These bottles are often designed to be able tohandle several times higher pressure in order to storehydrogen safely. Due to the weights of both the metalhydride and the bottles (typically made of stainless steel),the mass of hydrogen stored is normally o1.5%, which isquite low. Aluminum bottles that are reinforced by car-bon fibers and glass fibers are much lighter than stainless-steel bottles, but they cost much more. An overpressurerelief valve that will open when the hydrogen pressure

DC DC

DC

DC node

Stack

Battery orcapacitor

Motor

Air

H2

Control

Control

H2 tank

Blower

Figure 1 Function block diagram of a hydrogen–air fuel cell(FC) system. DC, direct current.

300

inside the bottle becomes higher than a predeterminedvalue is normally used for enhanced safety. A hydrogensensor may be used to detect hydrogen leakage. If ahydrogen leak is detected, ventilation should be activatedautomatically and the hydrogen supply line should beshut down. Contrary to public perception, hydrogen issafer than other fuels when there is a leakage becausehydrogen dissipates upward quickly due to its extremelylow density.

It should be understood that not all the hydrogenstored in a metal hydride can be released. It is anendothermic process where hydrogen is released, and thebottle can become quite cold, which also limits theamount of hydrogen that can be released. Air passing bythe bottle can be cooled down and then used for FC stackand system cooling. In contrast, fueling a metal hydridebottle with hydrogen is an exothermic process and thebottle becomes quite hot. This can limit the hydrogenfueling rate, and cooling the bottle is often necessaryduring the fueling process. Metal hydrides can be veryreactive with water, so it is important to prevent anywater from entering the bottle. In addition, the hydrogenstorage capacity of a metal hydride goes down with thenumber of charging/discharging cycles, and it needs tobe replaced with new ones after its capacity becomeslower than a desired value.

It is well known that hydrogen can cause brittleness ofmany metals because it can diffuse into the structures ofthese metals. So selecting the right cylinder material iscrucial for safe hydrogen storage. During a hydrogenfueling process, a hydrogen dispenser that transfershydrogen from a storage tank to a vehicle includes a hoseand a nozzle, and the nozzle must be sealed onto thevehicle fuel tank opening. A control system automaticallystops hydrogen dispensing when the tank is full.

A control unit (e.g., a microprocessor) coordinates theoperation of the entire system. It regulates the reactantflow rates to the stack according to the load demand fromthe FC. It monitors the condition of the FC stack and thebattery or capacitor, and displays it on a panel that can beeasily seen by a driver. It will shut down the FC systemwhen it detects problems such as stack overheating orhydrogen leakage.

Hydrogen flow to the stack should be either dead-ended or recirculated in order to minimize waste andavoid contaminating the environment. In a dead-endedconfiguration where the hydrogen outlet port is closedfor most of the time, hydrogen cannot be humidified, andperiodic opening of the hydrogen outlet port is oftennecessary to vent out accumulated water and any inertgases, such as nitrogen, that accumulate in the anodecompartment. Accumulation of water and inert gases inthe anode compartment can seriously lower an FC’sperformance in a dead-ended hydrogen configuration.This venting process is often called purging, because the

input hydrogen flow rate is normally increased in orderto purge water and inert gas out of the anode compart-ment in a short time (e.g., a fraction of a second). Somehydrogen also rushes out of the stack during purging tocause waste of fuel and contamination of the environ-ment. Ideally, this hydrogen should be collected for theFC to use, or it should be converted into water in ananode tail gas oxidizer or burner, but these processes maymake the system more complicated and bulkier. Thepurging frequency can be either programmed to occur atpredetermined time intervals or based on need. Theformer requires a thorough study and understanding ofthe FC behavior, and the FC behavior is predictable. Thelatter can be performed through the control algorithmwhen the stack voltage reaches a predetermined lowvalue.

A PEMFC can run longer and performs better whenthe membrane–electrode assemblies (MEAs) are prop-erly hydrated. Generally, reactants are humidified byhumidifiers before they enter the stack in order to hy-drate the MEAs. For an FC stack that generates 1 kW orgreater power, reactant humidification normally cannotbe avoided, which will in turn make the FC system morecomplicated, costly, and bulkier. For a stack generatingo1 kW of power, eliminating reactant humidification isfeasible. For example, the flow-field channels, shown inFigure 2, that were used at H Power (acquired by PlugPower Inc. in 2002) for nonhumidification stack operationconsist of dual inlets and outlets in an arrangement of onechannel’s inlet being placed near the other channel’soutlet so that the wet reactant in one channel shares itsmoisture with the drier reactant in the neighboringchannel to achieve better water uniformity in the entireactive area.

Related to MEA hydration is stack cooling. Besideselectricity, heat is also generated by an FC reaction, andthis heat needs to be removed in order to avoid MEAoverheating. A higher temperature will evaporate morewater from the MEA to make it drier. For a stack gen-erating a power of 1 kW or higher, a liquid coolant isoften needed. The coolant flows along the coolant flowfields made on the surface of coolant plates to remove theheat generated by the FC. Every cell or every severalcells can have a coolant plate. When a liquid coolant isused, a coolant tank, pump, meter, and valve will beneeded, in turn making the FC system more complicated.In addition, the coolant needs to be of negligible ionicconductivity, otherwise it will short the stack to causesevere damage; thus, a coolant purifier may also beneeded.

For a stack generating o1 kW power, air cooling maybe adequate. Air can be driven by fans or blowers. Foreffective heat removal, it is necessary to increase thestack’s surface-to-volume ratio. The ratio can be in-creased by either having transverse channels between

Applications – Transportation | Light Traction: Fuel Cells 301

cells (Figure 3) or having fins at the edge of the plates(Figure 4). For very low power stacks, blowing air overthe outside surface of the entire stack may be satisfactory.

Direct-Feed Liquid Systems

Due to the low hydrogen storage capacity and the lack ofhydrogen infrastructure, FCs using direct-feed liquidfuels are very attractive for light traction applicationsowing to their high volumetric energy density and ease ofstorage and transportation. The most widely investigatedsystem is the direct methanol fuel cell (DMFC).Methanol has an energy density of 6.3 kWh kg�1 or5.0 kWh L�1 based on the high heating value, but it oftenneeds to be highly diluted before it is fed to a stack.

Figure 5 shows a schematic functional diagram of aDMFC system. Methanol stored in a fuel tank is mixedwith water and the resulting mixture is sent to an FCstack. The methanol concentration in the mixing tank ismonitored by a methanol sensor. Dilute methanol and airundergo oxidation and reduction at the FC anode andcathode, respectively, to produce DC electrical power. Theunreacted methanol mixture is circulated back to the fuelmixing tank and carbon dioxide is vented out. Air exhaustis normally fully saturated with water, because a lot ofwater transports through the membrane from the anode to

the cathode. The water is condensed and collected intoeither a water tank or the fuel mixing tank.

For a DMFC system, the methanol concentration canhardly be 43 mol L�1 (B10wt%) for the followingreasons. First, because methanol has a relatively highpermeation rate through a membrane such as Nafion, andmethanol oxidation at the cathode can significantly lowerthe cathode potential, a higher methanol concentrationresults in a larger amount of methanol transporting to thecathode side and leads to higher cathode overpotentialloss and more waste of fuel. Second, the oxidation of eachmethanol molecule produces six protons, and each pro-ton can drag up to three water molecules from the anodeto the cathode. This means that 3 mol of methanol re-quires the presence of 54 mol of water, which translates toa maximum methanol concentration of about 3 mol L�1.The best FC performance is often achieved when about1 mol L�1 methanol solution is used.

Due to the permeation of methanol and its subsequentoxidation at the cathode, a DMFC often has an open-circuit voltage of only around 0.7 V. Because of the sig-nificant overpotentials suffered by both the anode and thecathode, it will be good performance if the cell voltagereaches 0.4 V at a current density of 0.2 A cm�2 at 60 1Ccell temperature. So, the power density of a DMFC stackis much lower than that of a PEMFC stack. Also, due to

Transversechannels

Figure 3 Illustration of air-cooled stacks with transversechannels.

Fins

Figure 4 Illustration of air-cooled stacks with fins.

Figure 2 Dual-inlet and dual-outlet flow-field design. Reprinted from Journal of Power Sources 109, Qi Z and Kaufman A, PEM fuelcell stacks operated under dry-reactant conditions, 469–476, Copyright (2002), with permission from Elsevier.

302 Applications – Transportation | Light Traction: Fuel Cells

the high methanol crossover rate, a significant portion ofthe fuel is wasted. Hydrocarbon membranes may havebetter ability to block methanol crossover than per-fluorinated membranes.

The MEAs in a DMFC are always fully hydratedbecause of the presence of a large quantity of water in theanode fuel mixture. Therefore, cathode reactant hu-midification is rarely required. Instead, dry air is pre-ferred in order to reduce the severity of cathode flooding.Very often a quite high stoichiometric air flow rate isneeded to minimize cathode flooding, but higher air flowrates make water balance more difficult. Cooling of the

cell is also not necessary due to the presence of the liquidfuel mixture.

Although methanol is the most widely investigatedliquid fuel for direct oxidation, its slow reaction kinetics,high crossover rate, and toxicity limit its applications. Ithas been shown that some other liquid fuels such as 2-propanol can perform better (Figure 6) before the anodeis seriously poisoned by the reaction intermediates.

Methanol can also be fed to an FC anode as vapor.The vapor can be formed through either a naturalevaporation of methanol at the cell temperature orboiling of methanol with a heat supply. This system,

DC DC

DC

DC node

Stack

Battery orcapacitor

Motor

Air

Control

Control

Fuelmixingtank

Methanoltank

Watertank

Methanol in

Methanol out

Figure 5 Function block diagram of a methanol/air fuel cell (FC) system. DC, direct current.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 50 100 150 200 250 300 350

Current density (mA cm−2)

Cel

l vol

tage

(V

)

1.0 mol L−1 methanol, air flow rate 920 mL min−1

1.0 mol L−1 methanol, air flow rate 180 mL min−1

1.0 mol L−1 2-propanol, air flow rate 920 mL min−1

1.0 mol L−1 2-propanol, air flow rate 180 mL min−1

Figure 6 Performance of 2-propanol vs methanol. Reprinted from Journal of Power Sources 112, Qi Z and Kaufman A Performance of2-propanol in direct-oxidation fuel cells, 121–129; Copyright (2002) with permission from Elsevier.


however, may be more complicated and may encounterhigher engineering challenges.

Light Traction Vehicles Powered By FuelCells

The research and development activity in this area ismuch lower than that for automobiles such as passengercars and commercial buses. One of the largest coopera-tive efforts so far is probably the HYCHAIN MINI-TRANS project that will deploy several fleets of in-novative FC vehicles in four regions of Europe operatingon hydrogen as an alternative fuel. It is expected thatlight traction vehicles such as scooters and electric-as-sisted bicycles will become more popular as oil pricescontinue to increase and possibly remain high, and as theenvironmental regulations become tougher. Most auto-mobile trips are o15 km and such a distance can be easilyhandled by scooters and bicycles. Also, it is really a wasteof energy when considering that about half of the tripsinvolve only a driver.

The major components of a conventional PEV arethrottle, battery, battery charger, motor speed controller,and electric DC motor. The throttle sends an electronicsignal to the motor controller to allow the rider to changeand control the speed. The battery provides the electricpower to the motor, and the battery charger is for char-ging the battery. The motor speed controller controls thepower to send to the motor based on inputs from thethrottle and the motor speed.

For an FC system-powered personal vehicle, thebattery and battery charger are replaced by an FC sys-tem. Because there is very limited knowledge in thepublic domain about FC-powered light traction vehicles,the following discussion is based on a generic study andsome examples.

Scooters

Scooters are two-wheel vehicles with a power require-ment somewhere between that of motorcycles and elec-tric bicycles. The driver rests his/her feet on a platformdirectly in front of him/her. The curb weight, which isthe vehicle weight without driver and cargo, should beo130 kg for off-power maneuverability. Highly pollutingtwo-stroke 50-cc ICE scooters are popular in Asiancountries. Scooters with four-stroke ICEs that can offersignificant pollution reduction should be adopted. Somecountries have mandated the adoption of zero-emissionelectric scooters. These electric scooters are powered bybatteries that need to be recharged every day for severalhours to overnight, and the range of travel after eachrecharging is very limited. The inconvenience of fre-quent recharging and short travel distance makes themless attractive to consumers than the ICE-powered

counterparts. This situation provides a good opportunityfor developing FC-powered scooters.

In 2000, Lin published an insightful comparison studybetween battery-powered and PEMFC-powered scootersbased on a standard road load model that is governed byeqns [1] and [2]:

Pwheels ¼ mav þ mgv sin yþ mgvCRRcos yþ 0:5pairCDAFv ½1�

Poutput ¼ Pwheels=Zdrivetrain þ Pauxiliary þ Pparasitic ½2�

where Pwheels is the mechanical power demand at wheels;m the combined mass of vehicle, driver, and cargo; a theacceleration rate; v the velocity of vehicle; g the gravi-tational acceleration rate; CRR the coefficient of therolling resistance; pair the density of air (1.23 kg m

�3); CDthe drag coefficient; AF the vehicle frontal area; Poutputthe engine power output; Zdrivetrain the efficiency ofelectric motor and controller subsystem (Lin used 77%);Pauxiliary the auxiliary power; and Pparasitic the parasiticpower needs. Auxiliary power units consist of data ac-quisition and panel display for parameters such as bat-tery/capacitor/FC voltage and current, scooter speed,elapsed time, fuel pressure (or amount), and tempera-tures. Equation [1] shows that heavier vehicles, highervelocity, higher acceleration rate, steeper slope, and lar-ger vehicle frontal area will result in higher power de-mand and shorter driving distance. Table 1 lists theperformance requirements for a mid-range powerscooter. The values in the brackets are Lin’s modelingresults for engine power output needs assuming a driverof 75 kg and an average auxiliary power of 60 W.

Lin used the Taipei Motorcycle Driving Cycle(TMDC) to calculate maximum power and average powerneeds for urban driving, and the results are summarized inTable 2. Clearly, due to the unavoidable need for accel-eration and hill climbing during routine driving, themaximum power required from the engine becomes aboutnine times as large as that required for steady-statecruising (5.6 kW vs 615 W). Also, the largest portion ofpower is consumed by acceleration. Further, the 5.6 kWhenergy requirements for 200 km is about four times theenergy amount stored by regular scooter batteries in 2000.

It is necessary to know that, in contrast to ICEs, if anelectric motor is not powerful enough to move a scooterwhen the throttle is turned to the maximum, the motoror controller will often burn out. Therefore, any FCsystem that is considered for powering the scooter mustbe able to generate a maximum power of 5.6 kW.

With this information, Lin then modeled PEMFC-powered scooters in three configurations: pure FC(5.9 kW), hybrid 1 with 3.2 kW FC and 2.6 kW battery,and hybrid 2 with 1.1 kW FC and 4.6 kW battery. A basicrule was that the battery provided supplemental powerwhenever the FC maximum power was insufficient for


the driving requirements; 300 W was added on top of the5.6 kW due to the parasitic power requirements of an FCfor powering pumps, fans, and blowers. In all three cases,hydrogen was supplied in a dead-ended configuration, airwas provided at a 2.5 times stoichiometric flow rate, li-quid cooling was employed, FeTiH1.9 metal hydride thatcould store hydrogen at 1.75wt% and 101 g L�1 was used,and state-of–the-art FC performance was assumed (0.82,0.71, and 0.61 V cell voltage at 200, 600, and1000 mA cm�2 current densities, respectively). Table 3summarizes the FC design specifications and the scooterperformance under TMDC. The model results indicatedthat, in all the three cases, the FC vehicles offered more

than three times the range of Taiwan ZES-200 electricscooters. The FC systems were only slightly heavier butrequired a lot more space due to the bulky auxiliaries,liquid cooling, and hydrogen storage canisters. The pureFC configuration had slightly higher fuel efficiency andbetter performance with respect to maximum powerendurance and 3.2 kW hill climbing endurance than thehybrid configurations. The inferior performance of thehybrid configurations was due to the use of the powercomplementary batteries. In addition, the FC-poweredscooters can have more than three times the fuel effi-ciency of gasoline-powered two-stroke ICE scooters.

Since 2000, tremendous progress has been made in thearea of FC technologies. For example, a prototypemotorbike developed by Intelligent Energy (London,England) used a 1-kW PEMFC stack with a powerdensity of 2.5 kW L�1. The stack did not require eitherliquid cooling or reactant humidification. The FCachieved full power almost instantaneously at roomtemperatures and in less than 60 s at –25 1C. Combinedwith a 720-Wh battery pack (4� 12 V� 15 Ah), themotorbike accelerated from zero to the top speed of80 km h�1 in 12.1 s using a 6-kW, 48-VDC brushedmotor. With 2.4 kWh hydrogen stored onboard in highpressure and lightweight reinforced aluminum cylinders,the motorbike that weighed 80 kg could travel at least160 km before refueling. The refueling took o5 min.

Figure 7 shows an FC-powered scooter developed byPalcan Power Systems Inc. (Canada) in 2003. The scooterused a 700-W electric motor powered by a1.5-kW PEMFC. The FC stack contained 40 cells andwas cooled by liquid water. Two batteries rated at12 V and 12 Ah were used to assist the FC during ac-celeration and hill climbing. Regenerative breaking wasused to recharge the battery during breaking. Threemetal hydride canisters with a total weight of 10.5 kgwere used to store hydrogen at a pressure of 690 kPa.Hydrogen was fed to the stack in a dead-ended con-figuration. The scooter could travel for 70 km after eachrefueling of 0.9 m3 hydrogen, whereas a battery-poweredcounterpart could only travel for about half of that dis-tance. It had a maximum speed of 50 km h�1 and couldclimb slopes up to 251 with a 70-kg rider.

When considering where to install an FC system on ascooter, the center of gravity and weight distributionshould be considered along with the availability of space.Because a lower center of gravity makes a scooter morestable, heavy components of an FC system should be ar-ranged as low as possible. Also, the components should bearranged in such a way that a proper weight distribution isachieved to avoid instability over bumps if the weight istoo far forward and improper front wheel tracking if theweight is too far backward. Further, the hydrogen storagebottles should be put in a strong protective case for en-hanced safety.

Table 2 Modeled power requirement for TMDC and steady-state driving

Specification TMDC 30 km h�1

cruising

Total drive time 950 s 950 s

Average speed 19.3 km h�1 30 km h�1

Maximum speed 46 km h�1 30 km h�1

Maximum acceleration 6.4 km h�1 s�1 NA

Maximum deceleration –8.6 km h�1 s�1 NA

Maximum power from

engine

5.6 kW 615 W

Average power from engine

(exclude parasitic)

566 W 615 W

Average acceleration power 215 W NA

Average rolling resistance

power

151 W 305 W

Average aerodynamic drag

power

117 W 250 W

Mileage per kilowatt-hour 35.5 km 48.8 km

Total energy requirement

for 200 km

5.6 kWh 4.1 kWh

NA, not applicable; TMDC, Taipei Motorcycle Driving Cycle.

Source: Reprinted from Journal of Power Sources 86, Lin B, Conceptual

design and modeling of a fuel cell scooter for urban Asia, 202–213,

Copyright (2002), with permission from Elsevier.

Table 1 Performance requirements for a mid-range powerscooter

Specification Requirement

Maximum motor power

output

4–6 kW

Travel range per refueling 200 km at 30 km h�1 cruising

velocity

Fuel efficiency 4100 mpgeAcceleration 0–30 km h�1 in less than 5 s

Speed on 151 slope 10 km h�1 (requires 2050 W)Speed on 121 slope 18 km h�1 (requires 3020 W)Maximum speed 60 km h�1 (requires 2600 W)

Curb weight o130 kgmpge, mile per gallon equivalent.

Source: Reprinted from Journal of Power Sources 86, Lin B,

Conceptual design and modeling of a fuel cell scooter for urban Asia,

202–213, Copyright (2002), with permission from Elsevier.


Bicycles

Bicycles are the most popular commuting vehicles for atravel distance o10 km in many Asian countries. In re-cent years, electric-assisted bicycles powered by batterieshave started picking up popularity. For an electric-

assisted bicycle, the US Federal Government regulatesthat the vehicle must have operable pedals, a motor of nomore than 750 W, and cannot be operated under power>32 km h�1. Because the rider can still pedal during ac-celeration and hill climbing, electric bicycles normally donot need motor power 4400 W. This makes their designmuch easier and their cost more affordable than scootersthat need about nine times as power output during ac-celeration and hill climbing as steady-state cruising.Lead–acid and nickel–cadmium batteries are widely usedto provide the electric power. Similar to scooters, due tothe inconvenience of frequent recharging and the lowenergy storage capacity of batteries, a PEMFC-poweredbicycle could be more advantageous.

Hwang and coworkers installed a 40-cell air-cooledstack that provided a nominal power of 303 W at 0.7 Vand a peak power of 378 W onto a 30-kg bicycle.Hydrogen stored in a metal hydride canister was suppliedto the stack at 48 kPa in a dead-ended configuration. Incombination with a pressure regulator, the hydrogencanister provided a passive and simple load-followingmechanism: a solenoid valve opened as the pressure inthe fuel feeding line decreased to a predetermined valuedue to the consumption of hydrogen at the anode. Ananode purging valve opened when the stack voltage

Table 3 Fuel cell (FC) system design specifications and scooter performance under TMDC

Specification Fuel cell Hybrid 1 Hybrid 2

Maximum fuel cell gross power 5.9 3.2 1.1

Current at maximum power (A) 172 89 31

Efficiency at maximum power (%) 41.2 43.2 43.6

Electrode active area (cm2) 170 100 35

Total active area needed (cm2) 9600 5600 2000

Maximum heat generation (kW) 7.0 3.5 1.2

Maximum fuel cell net power (kW) 5.9 3.0 1.0

Assisting battery power (kW) 0 2.6 4.6

Cooling fan power need at steady 3020 W (W) 14 28 4

Coolant pump power need at steady 3020 W (W) 25 38 21

Average total power output (W) 674 709 726

Average fuel cell power (W) 674 698 577

On-vehicle fuel economy (mpge) 344 316 343

Fuel cell conversion efficiency (%) 56 53 47

Average power regenerated through braking (W) 0 65 86

Braking energy recovered/maximum braking loss 0 0.51 0.68

Fuel cell stack weight (kg)/volume (L) 7.6/7.8 5.4/5.3 4.1/3.2

Auxiliary weight (kg)/volume (L) 17.5/22.0 15.6/21.3 10.5/8.4

Battery weight (kg)/volume (L) 0/0 3.1/3.0 5.6/5.4

Stack and combined auxiliary weight (kg)/volume (L) 25.1/29.8 24.6/29.6 20.2/17.0

Motor and controller weight (kg)/volume (L) 15.5/9.1 15.5/9.1 15.5/9.1

Hydride canister for 200 km weight (kg)/volume (L) 21.4/3.7 22.8/4.0 24.3/4.2

H2 mass for 200 km range (g) 250 266 285

Total drive system weight (kg)/volume (L) 61/43 63/43 60/30

Duration to sustain maximum power (s) Unlimited B10 B10Duration to sustain 3.2 kW hill climbing (min) Unlimited Unlimited o3mpge, mile per gallon equivalent; TMDC, Taipei Motorcycle Driving Cycle.

Source: Reprinted from Journal of Power Sources 86, Lin B, Conceptual design and modeling of a fuel cell scooter for urban Asia, 202–213,

Copyright (2002), with permission from Elsevier.

Figure 7 Picture of a scooter. Courtesy of Palcan PowerSystems Inc.


dropped to a predetermined value of 22 V. Cooling fanswere turned on when the stack temperature was 440 1C.Neither hydrogen nor air was humidified. During a 1-hroller-stand test, the bicycle reached a maximum speedof 25.2 km h�1 with reliable operation. During a road testwith a 70-kg rider but without pedaling, the bicyclereached a maximum speed of 16.8 km h�1, and 6.8 ghydrogen was sufficient for a driving distance of 9.2 km,i.e., the bicycle had a distance-to-fuel ratio of 1.35 km pergram of hydrogen. The system efficiency varied fromabout 24% at 100 W power output to 34% at 315 Wpower output. In 2004, an FC-powered bicycle jointlydeveloped by US and Italian inventors reached a topspeed of 32 km h�1 with no pedaling and traveled 100 kmwith a small tank of hydrogen.

Forklifts

Battery-powered forklifts are used in warehousesworldwide. It is estimated that there are about 2.5 millionforklifts in North America and Europe. The battery hasto be either replaced or recharged after running foro8 h, and the lifting may become sluggish even beforethe battery replacement or recharging. Fuel cell/battery(or supercapacitor) powered forklifts developed andtested by several companies have shown much betterperformance, and the total package can be much smallerand lighter than a pure battery system. For example, a2300-kg forklift that was powered by a 12-kW HyPm 12-FC stack in conjunction with a supercapacitor developedby General Motors (GM) of Canada and HydrogenicsCorp. ran 12 h before the hydrogen tank (containing1.8 kg H2) was refilled, and the refueling only took 2 min.In addition, without any modification of current forklifts,there is more than enough space to host an FC system ifthe batteries are replaced. One of GM’s onsite electro-lyzers, which generated 65 kg of hydrogen at 34.5 MPaper day, could serve 25 forklifts for three shifts daily. Ittook up less floor space than the charger and batterysystem, which could only supply three to four batteriesdaily.

Other Light Traction Vehicles

FCs are expected to be in a more advantageous positionthan batteries to power other light traction vehicles thattravel in the sky, on the ground or water, or in the ocean.For example, in June 2004, Asia Pacific Fuel CellTechnologies, Ltd (Taiwan) completed a second-gener-ation wheelchair that was powered by a PEMFC andnickel–metal hydride (Ni–MH) battery with a combinedmaximum power of 1000 W. The wheelchair reached amaximum speed of 6 km h�1, and it could continuouslyoperate for 12 h for 60 km. An AB5-type metal hydridewas used to store hydrogen in extruded aluminum alloybottles. A bottle with a diameter of 76.2 mm and length of

365 mm stored 0.54 m3 hydrogen at 30 1C, and the totalcanister weighed about 4.5 kg. Figure 8 shows a wheel-chair-type personal mobility vehicle developed by Bei-jing Fuyuan Century Fuel Cell (China).

FCs have also been demonstrated for powering golfcarts and tour carts. Figure 9 shows an FC engine–powered four-seat tour cart developed by ShanghaiShen-Li High Tech Co., Ltd (China) in 2004. The FCengine weighed 100 kg and had a power output of 5 kW,with an electrical efficiency 450%. The cart demon-strated a travel range of 200 km with the amount ofhydrogen stored in a reinforced aluminum bottle. Thebottle had a volume of 0.07 m3, weight of 40 kg, and thehydrogen was pressurized at 20 MPa. The cart could

Figure 8 Picture of a wheelchair. Courtesy of Beijing FuyuanCentury Fuel Cell.

Figure 9 Picture of a tour cart. Courtesy of Shanghai Shen-LiHigh Tech.


reach a top speed of 20 km h�1 in 30 s. The cart wasexclusively powered by the FC while a very small batterywas used to start the system. This cart achieved a total of20 000 km driving distance as of July 2006.

Illustrated in Figure 10 is a mini scooter on which anoperator will stand that was developed by GasHubTechnology (Singapore). The scooter was powered by a200-W air-cooled PEMFC in conjunction with twoultracapacitors (58 F, 15 V each) that assisted accelerationand hill climbing, and it reached a top speed of 20 km h�1

with a travel range between 50 and 80 km. The netweight of the scooter was 23 kg, its load-carrying capacitywas 100 kg, and it had four speed levels and could climb aslope up to 101. The two metal hydride cylindersweighed 6 kg and contained a total of 0.6 m3 of hydrogen.In comparison, a similar scooter powered by batteriestraveled for only 20–30 km.

Challenges

There are still many hurdles to overcome before FC-powered light traction vehicles are widely accepted bythe general public. First, the cost of FC systems has tocome down so that they are affordable and their life cyclecost is comparable to conventional power sources. Sec-ond, the reliability and durability of FC systems have toimprove to be as good as current power sources. Third, ahydrogen infrastructure that includes hydrogen pro-duction, transportation, storage, and fueling needs to beestablished, and the hydrogen mass storage capacity mustbe increased. As of August 2006, there were about 140hydrogen fueling stations worldwide, which is too few tobe of any significance. Figure 11 shows China’s firsthydrogen generation and fueling station of its type that

was built by Beijing LN Power Sources Co. Ltd in 2006.This station could produce 300 Nm3 h�1 hydrogenthrough electrolysis of water (Figure 12), and thehydrogen is compressed to 40 MPa. The station is used torefuel not only automobiles but also hydrogen bottles andoxygen bottles. Fourth, as conventional power sourcesbecome cleaner and more efficient, it will be eventougher for FCs to compete. Fifth, the misperception thathydrogen is more dangerous than other fuels must becorrected through public education so that the generalpublic realizes that hydrogen is just as safe as othergaseous or liquid fuels when handled properly.

Figure 10 Picture of a mini scooter. Courtesy of GasHubTechnology.

Figure 11 Picture of a hydrogen generation and refueling station. Courtesy of Beijing LN Power Sources.


Conclusions

The feasibility of FC-powered light traction vehicles hasbeen demonstrated by a number of companies worldwide.Fuel cell or FC/battery (or supercapacitor) hybrid sys-tems enable such vehicles to operate several times longerthan a battery system. The FC or its hybrid system canalso be several times more efficient than ICEs. Hydrogenstored in either metal hydride cylinders or high-pressurebottles is almost exclusively used.

Nomenclature

Symbols and Units

a acceleration rate

AF vechicle frontal area

CD drag coefficient

CRR coefficient of rolling resistance

g gravitational acceleration rate

m combined mass of vehicle, driver, and

cargo

pair density of air

Pauxiliary auxiliary power

Poutput engine power output

Pparasite parasitic power needs

Pwheels mechanical power demand at wheels

v velocity of vehicle

gdrivetrain efficiency of electric motor and con-

troller subsystem

Abbreviations and Acronyms

DC direct current

DMFC direct methanol fuel cell

FC fuel cell

GM General Motors

ICE internal combustion engine

MEA membrane-electrode assembly

mpge mile per gallon equivalent

PEMFC proton-exchange membrane fuel cell

PEV personal electric vehicle

TMDC Taipei Motorcycle Driving Cycle

Further Reading

Beckhaus P, Dokupil M, Heinzel A, Souzani S, and Spitta C (2005) On-board fuel cell power supply for sailing yachts. Journal of PowerSources 145: 639--643.

Büchi FN (2003) Small-size PEM systems for special applications.In: Vielstich W, Lamm A, and Gasteiger HA (eds.) Handbook of FuelCells – Fundamentals Technology and Applications, vol. 4,pp. 1152--1161. Chichester: John Wiley & Sons Ltd.

Chalkova E, Pague MB, Fedkin MV, Wesolowski DJ, and Lvov SN(2005) Nafion/TiO2 proton conductive composite membranes for

Figure 12 Picture of an electrolysis system. Courtesy of Beijing LN Power Sources.


PEMFCs operating at elevated temperature and reduced relativehumidity. Journal of the Electrochemical Society 152:A1035--A1040.

Chen CY and Yang P (2003) Performance of an air-breathing directmethanol fuel cell. Journal of Power Sources 123: 37--42.

Chow CY and Wozniczka BM (1995) Electrochemical Fuel Cell Stackwith Humidification Section Located Upstream from theElectrochemically Active Section. US Patent 5,382,478, 17 January.

Colella WG (2000) Market prospects, design features, and performanceof a fuel cell-powered scooter. Journal of Power Sources 86:255--260.

Conway BE (1999) Electrochemical Supercapacitors – ScientificFundamentals and Technological Applications. New York: PlenumPublishers.

Dong Q, Mench MM, Cleghorn S, and Beuscher U (2005) Distributedperformance of polymer electrolyte fuel cells under low-humidityconditions. Journal of the Electrochemical Society 152:A2114--A2122.

Funck R (2003) High pressure storage. In: Vielstich W, Lamm A, andGasteiger HA (eds.) Handbook of Fuel Cells – FundamentalsTechnology and Applications, vol. 3, pp. 83--88. Chichester: JohnWiley & Sons Ltd.

Gilman S and Chu D (2003) Methanol effects on the O2 reductionreaction. In: Vielstich W, Lamm A, and Gasteiger HA (eds.)Handbook of Fuel Cells – Fundamentals Technology andApplications, vol. 2, pp. 652--661. Chichester: John Wiley & SonsLtd.

Hamnett A (2003) Direct methanol fuel cells (DMFC). In: Vielstich W,Lamm A, and Gasteiger HA (eds.) Handbook of Fuel Cells –Fundamentals Technology and Applications, vol. 1, pp. 305--322.Chichester: John Wiley & Sons Ltd.

Hwang JJ, Wang DY, and Shih NC (2005) Development of a lightweightfuel cell vehicle. Journal of Power Sources 141: 108--115.

Hwang JJ, Wang DY, Shih NC, Lai DY, and Chen CK (2004)Development of fuel-cell-powered electric bicycle. Journal of PowerSources 133: 223--228.

Jones DJ and Rozière J (2003) Inorganic/organic compositemembranes. In: Vielstich W, Lamm A, and Gasteiger HA (eds.)Handbook of Fuel Cells – Fundamentals Technology andApplications, vol. 3, pp. 447--455. Chichester: John Wiley & SonsLtd.

Kocha SS (2003) Principles of MEA preparation. In: Vielstich W, LammA, and Gasteiger HA (eds.) Handbook of Fuel Cells – FundamentalsTechnology and Applications, vol. 3, pp. 538--565. Chichester: JohnWiley & Sons Ltd.

Lamm A and Müller J (2003) System design for transport applications.In: Vielstich W, Lamm A, and Gasteiger HA (eds.) Handbook of FuelCells – Fundamentals Technology and Applications, vol. 4,pp. 878--893. Chichester: John Wiley & Sons Ltd.

Larminie J and Dicks A (2003) Fuel Cell Systems Explained, 2nd edn.New York: John Wiley & Sons, Ltd.

Lin B (2000) Conceptual design and modeling of a fuel cell scooter forurban Asia. Journal of Power Sources 86: 202--213.

Lin J-C, Kunz HR, and Fenton JM (2003) Membrane/electrodeadditives for low-humidification operation. In: Vielstich W, Lamm A,and Gasteiger HA (eds.) Handbook of Fuel Cells – FundamentalsTechnology and Applications, vol. 3, pp. 456--463. Chichester: JohnWiley & Sons Ltd.

Lipman T and Sperling D (2003) Market concepts, competingtechnologies, and cost challenges for automotive and stationaryapplications. In: Vielstich W, Lamm A, and Gasteiger HA (eds.)Handbook of Fuel Cells – Fundamentals Technology andApplications, vol. 4, pp. 1318--1328. Chichester: John Wiley & SonsLtd.

Mathias MF, Roth J, Fleming J, and Lehnert W (2003) Diffusion mediamaterials and characterisation. In: Vielstich W, Lamm A, andGasteiger HA (eds.) Handbook of Fuel Cells – FundamentalsTechnology and Applications, vol. 3, pp. 517--537. Chichester: JohnWiley & Sons Ltd.

Qi Z, He C, Hollett M, Attia A, and Kaufman A (2003) A reliable and fast-responding methanol concentration sensor with novel design.Electrochemical and Solid-State Letters 6: A88--A90.

Qi Z and Kaufman A (2002a) Improvement of water management by amicroporous sublayer for PEM fuel cells. Journal of Power Sources109: 38--46.

Qi Z and Kaufman A (2002b) PEM fuel cell stacks operated under dry-reactant conditions. Journal of Power Sources 109: 469--476.

Qi Z and Kaufman A (2002c) Performance of 2-propanol in direct-oxidation fuel cells. Journal of Power Sources 112: 121--129.

Sandrock G (2003) Hydride storage. In: Vielstich W, Lamm A, andGasteiger HA (eds.) Handbook of Fuel Cells – FundamentalsTechnology and Applications, vol. 3, pp. 101--112. Chichester: JohnWiley & Sons Ltd.

Scheffler G, Wurster R, and Schindler J (2003) Hydrogen safety, codesand standards for vehicles and stationary applications. In: VielstichW, Lamm A, and Gasteiger HA (eds.) Handbook of Fuel Cells –Fundamentals Technology and Applications, vol. 3, pp. 257--267.Chichester: John Wiley & Sons Ltd.

Uchida H, Ueno Y, Hagihara H, and Watanabe M (2003) Self-humidifying electrolyte membranes for fuel cells – preparation ofhighly dispersed TiO2 particles in Nafion 112. Journal of theElectrochemical Society 150: A57--A62.

Wind J, LaCroix A, and Braeuninger et al. (2003) Metal bipolar platesand coatings. In: Vielstich W, Lamm A, and Gasteiger HA (eds.)Handbook of Fuel Cells – Fundamentals Technology andApplications, vol. 3, pp. 294–307. Chichester: John Wiley & SonsLtd.

Wolf J (2003) Liquid hydrogen technology for vehicles. In: Vielstich W,Lamm A, and Gasteiger HA (eds.) Handbook of Fuel Cells –Fundamentals Technology and Applications, vol. 3, pp. 89--100.Chichester: John Wiley & Sons Ltd.


Outline placeholderIntroductionFuel Cell SystemsHydrogen-Air SystemsDirect-Feed Liquid Systems

Light Traction Vehicles Powered By Fuel CellsScootersBicyclesForkliftsOther Light Traction VehiclesChallenges

ConclusionsFurther Reading

light traction: fuel cells - elsevier.com · 2013. 12. 18. · light traction applications, it is...

Documents